|Publication number||US6340903 B1|
|Application number||US 09/567,948|
|Publication date||Jan 22, 2002|
|Filing date||May 10, 2000|
|Priority date||May 10, 2000|
|Publication number||09567948, 567948, US 6340903 B1, US 6340903B1, US-B1-6340903, US6340903 B1, US6340903B1|
|Inventors||James W. Leith|
|Original Assignee||Zilog, Ind.|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (15), Non-Patent Citations (14), Referenced by (7), Classifications (5), Legal Events (6)|
|External Links: USPTO, USPTO Assignment, Espacenet|
1. Field of the Invention
The present invention relates generally to sample and hold circuits and, more particularly, to a method and apparatus for extracting a time varying component of an analog signal while resetting the DC level of the signal to a new reference value.
2. Background Information
Sample and hold (S/H) stages are common in many analog circuits, for example in analog to digital converters and other applications where an analog signal needs to be sampled and held for some period of time. These circuits generally function by having a switch through which the input voltage which is to be sampled is supplied to the S/H circuit and a capacitor on which this sampled charge is stored. In the sample mode, the circuit output is derived from the input and the capacitor is charged. During the usually much longer hold mode, the output is based upon the changed placed on the capacitor by the input voltage during the brief sample mode.
An example of a prior art design is shown in FIG. 1a. In this particular design, the charge holding capacitor Chold 24 is part of a feedback loop to the same op-amp input that receives the input signal Vin, although a number of other placements can be found in other designs. This particular arrangement is an integrator-type S/H circuit. When switch S1 25 is closed and S2 26 is open, the circuit is in sample or track mode and charge accumulates on Chold 24. The equivalent circuit is shown in FIG. 1b. The DC gain is given by the ration of Rf 22 to Ri 23 and the overall transfer function is Vout=−(Rf/Ri)(1sRfChold)−1Vin, where s is the Laplace transform parameter, so that it acts as an inverting low-pass circuit. The charge accumulated on Chold 24 is then proportional to Vout=−(Rf/Ri)(Vin−V+), where V+ is the voltage at the non-inverting terminal of op-amp 21.
The equivalent circuit during hold mode is shown in FIG. 1c. S1 25 is open and S2 26 is closed, and Vout is determined by the integrated charge stored on Chold 24 during the sample mode. S2 26 minimizes signal feedthrough when in hold mode and keeps the common node of the resistive network close to the voltage required when the circuit goes back to the sample mode.
One application of a sample and hold circuit is to extract a time varying component of an analog signal having a DC component of much greater amplitude than the time varying component. The circuit will need to pass and amplify the frequencies in the range of interest while at the same time largely removing any DC component. For example, if the circuit needs to extract a signal with a bandwidth of 10-1000 Hz embedded in an analog signal with a DC level several orders of magnitude larger this time varying portion, a S/H circuit as in FIG. 1 is inadequate: As a low pass filter, it will amplify the DC component (as well as any offset voltages passed on) more that the signal in the desired bandwidth; and if the resistors are chosen so that this bandwidth is amplified enough to be easily detectable, the DC level will have saturated the output thereby destroying any time varying component. Additionally, its low pass function makes it difficult to obtain a fast settling speed within acceptable values for the resistors and capacitor.
These problems can be reduced by giving the S/H circuit itself a high gain high pass filter function, combining it with additional stages high gain high pass sections, or both. The standard of doing this is to use a large capacitance either in place of, or in series with, the resistor Ri 23 in FIG. 1a. If an additional amplification stage with a standard, non-S/H high pass high gain filter function is added, it would be similar to FIG. 1 a with the integrating capacitor Chold 24 and the switches S1 25 and S2 26 removed, but a large capacitor would still be needed in the gain path. Although this sort of arrangement may give the desired response characteristics, it produces a number of new problems, largely related to the need for the large electrolytic capacitors in the gain path of the circuit. Additionally, when a gain of several orders of magnitude is needed for the wanted signal, any stray offset voltages are similarly magnified.
Having large capacitors causes several problems as such capacitors are not readily incorporated into an integrated circuit. Prior art solutions thus involve using discreet S/H circuits, resistors/capacitors, and operational amplifiers. This results in the reliability problems associated with discreet solutions, such as board leakage around the (usually FET) S/H and failure of the large electrolytic AC coupling capacitors in the signal path. Use of discreet components also limits the reduction both of the number of components and of the required board area. Additionally, these large off-chip capacitors have problems in radiation due to a susceptibility to induced electromagnetic fields.
The present invention provides a sample and hold circuit which uses an auto-zero feedback technique to cancel the DC level of the input signal and reference this signal to a new baseline. The circuit is based on an op-amp with two separate feedback loops. A first feedback loop is connected to the same op-amp input as the incoming signal and contains a capacitor to store charge from this signal during sample mode and set the output voltage during hold mode. The second feedback loop uses an auto-zero feedback technique and contains an integrator having a predetermined reference voltage, thereby allowing the DC level of the input signal to removed without the need for capacitors in the gain path of the circuit. This allows the sample and hold circuit to extract an embedded time varying signal from the input voltage. It can be configured for a high gain, high pass function, again without the need for large electrolytic capacitors in the gain path, removing the problems associated with such capacitors.
An exemplary embodiment is as a three stage circuit to extract an embedded signal lying in a frequency range, but having an amplitude much smaller than the DC component of the analog signal within which it is embedded. An initial low gain section acts as a buffer before the sample and hold section. Following the sample and hold section, the circuit employs a stage with a high gain high pass function also using auto-zero feedback. This again eliminates the large electrolytic capacitors usually placed in the gain path that can not be implemented as part of a single integrated circuit. Besides eliminating the need for off-chip capacitors in the gain path, by using auto-zero feedback all sample and hold errors are referenced to the output of the sample and hold stage and are not amplified by the total system in a multi-stage configuration as the offset voltages of the amplifiers are not amplified.
Additional objects, advantages, and features of the present invention will become apparent from the following description of its preferred embodiments, which description should be taken in conjunction with the accompanying drawings.
FIG. 1a is an example of one type of prior art sample and hold circuit.
FIGS. 1b and 1 c are equivalent circuits of FIG. 1a in, respectively, sample and hold mode.
FIG. 2 is a multi-stage sample and hold circuit embodying the present invention.
FIGS. 3a, 3 b, and 3 c show exemplary waveforms for Vsam, Vin, and Vout respectively.
FIG. 2 is a multistage sample and hold circuit embodying many aspects the present invention. This example is configured with three stages. The first stage is a non-inverting amplifier Ain 110 to receive the input voltage Vin. The second stage is the sample and hold section S/H 120. The last stage is the amplifier circuit Aout 150. Both the sample and hold section S/H 120 and the subsequent stage Aout 150 receive the input from the preceding section through the inverting inputs of their respective op-amps 121 and 151, both of which are also connected in a feed back loop. Both S/H 120 and Aout 150 have an additional feedback loop with an integrator, Ifs 140 and Ifa 160 respectively, into the non-inverting input of their respective op-amps. As described below, the inclusion of this second, auto-zero feedback loop cancels the DC level of the preceding sections and references the signal to a new baseline. As arranged in FIG. 2, this circuit can extract a small time varying component from an analog signal having a much larger constant voltage.
First, circuit Aout 150 is described, as it is similar to the sample and hold section S/H 120 but simpler. A discussion of the sample and hold section is then given, followed by the operation of the circuit as a whole.
In Aout 150, op-amp 151 receives the output of the preceding stage S/H 120 into the inverting output through resistor Ri3 153. Feedback resistor Rf3 152 connects this input to the output and the gain is proportional to the ratio of these two resistors. A second feedback loop is connected back to the second, non-inverting input of op-amp 151. This second feedback loop is based on an integrator Ifa 160 with a predetermined reference voltage Vref2. The integrator Ifa 160 is composed of a second op-amp 161 whose inverting input is connected to the output of the first op-amp 151 through resistor Raz2 163, with capacitor Caz2 162 in the feedback loop to provide the integration function. The predetermined reference voltage Vref2 is supplied to the non-inverting input and the output goes back to the non-inverting input of op-amp 151 to close the second feedback loop. The reference voltage Vref2 can either be fixed or adjustable, in either case by a resistive voltage divider or other standard technique. As discussed below, a convenient choice for the present application is Vref2=1/2Vdd.
The transfer function of Aout 150 is given by
where α=RazCaz, β=αRi/(Ri+Rf), s is the Laplace transform parameter, and the minus sign is due to the inverting gain topology used. Due to the first term in Vout, Aout 150 provides a high gain, high pass function. As both it and the last term go to zero for s=0, any DC offset from the previous stage is removed. The second term then references the output to a new base line of Vref2. Note that in addition to removing any DC component from the previous stage, it can also cancel any other offset voltages that would otherwise be introduced.
The use of the second feed back loop serves several purposes. First, it turns Aout 150 into a high pass filter since the expression in brackets goes to zero as sRaz2Caz2 for low frequencies and becomes (Ri3+Rf3)/Ri3 in the high frequency limit. Without this second feed back loop, the standard technique to obtain a high pass function is to place a large electrolytic capacitor in the gain path, either replacing, or in series with, the input resistor Ri3 153. As discussed in the Background section, the use of such a capacitor has several disadvantages. Second, it supplies “auto-zero feedback”, canceling the DC level of the previous section and referencing the signal to the new baseline of Vref2. Any offset voltages from the sample and hold section are thus eliminated instead of being amplified. Here, “auto-zero feedback” refers to setting the zero frequency level, which may or may not be ground since this level is set to zero only for the choice of Vref2=0. Additionally, as the standard frequency independent gain supplied through the first feed back loop, Rf3/Ri3, multiplies Vin, this allows the overall gain of section Aout 150 to be adjusted largely independently of both the characteristics of the filter function in the brackets and the new baseline level.
The middle stage in FIG. 2 is the sample and hold section S/H 120. This stage has a geometry similar to subsequent stage Aout 150. The input from the preceding stage is again received into the inverting input of op-amp 121, but now through switch S1 125, and capacitor Chold 124 is in the first feedback loop into this input. The second, auto-zero feedback loop again contains an integrator Ifs 140 with offset voltage Vref1.
By placing capacitor Chold 124 in the feedback loop, stage S/H 120 functions as an integrator type sample and hold circuit. When switch S1 125 is closed during track mode, the circuit acts as a filter and accumulates charge on Chold 124. In hold mode, S1 125 is open and the voltage at the inverting input is proportional to the integrated charge on Chold 124 accumulated while S1 125 was open. As with the circuit of FIG. 1a, Qhold is again proportional (−Rf2/Ri2)(VinS/H−V+S/H), where in this expression VinS/H is input voltage to S/H 120 through Ri2 and V+S/H is the voltage into the non-inverting input of op-amp 121; but, as the non-inverting input is now from Ifs 140, the difference of these two values in Qhold again re-references the DC level during the hold mode. the switch S1 125 is controlled by the sampling signal, here denoted as Vsam, which typically has a hold phase comprising only a small portion of its total period.
The resistor Rf2 122 is placed in parallel with Chold 124 in the first feedback loop, but connected with switch S1 125 in the path back to the input of op-amp 121. Including this resistor in the loop allows the gain of the sample and hold stage to be set and alters the transfer function of the stage. As some applications of sample and hold circuits require a high gain and particular frequency pass characteristics, this feedback resistor is used to set these properties, as discussed in the example below. Switch S2 126 is connected to ground, so that it can be closed during hold mode to minimize signal feed through. Both of switches S1 125 and S2 126 are controlled by the signal Vsam, S1 125 directly and S2 126 through an inverter, so that they function in unison by having S1 125 closed and S2 126 open during sample mode and the other way around during hold mode which usually comprises the much greater part of Vsam's period. Finally, the signal from the previous stage passes through an input resistor Ri2 123 before reaching these switches.
In sample mode, when switch S1 125 is closed and switch S2 126 is open, section S/H 120 will have the same transfer function as given above for Aout 150 in equation (A), but with Rf replaced by the parallel impedance of Rf2 122 and Chold 124: Rf2→Rf2(1+sRf2Chold)−1. Although the inverting feedback loop is the same as the circuit of FIG. 1a, the transfer function is no longer that of a low pass filter due to the [sα/(sβ+1)] factor now multiplying Vin from the lower, auto-zero feedback loop. As this term again goes to zero for zero frequency, the DC component is again removed. The second term in (A) again references the signal to a new baseline value of Vref1. The lower feedback loop with integrator Ifs 140 therefore functions as in section Aout 150, but has several additional advantages when incorporated into a sample and hold section as done here. In some prior art designs, integrators have been included in a feedback loop to alter the DC level of the output signal, but the hold capacitor has been part of this same feedback loop. By splitting the functions into separate feedback loops, one for the auto-zero function and one for the hold capacitor, the circuit can be made more flexible in addition to executing this individual functions better by adjusting the independent parameters in (A): Vref1, α1=Caz1Raz1, (Rf2Ri2), and CholdRf2.
In the embodiment of FIG. 2, the sample and hold stage S/H 120 receives its input from an initial amplification stage Ain 110. This first section is composed of an op-amp 111 that receives the initial circuit input Vin into its non-inverting input. The inverting input is shown connected to ground through a resistor Ri1 113 and is also connected in a feedback loop through resistor Rf1 112. The gain of this stage is therefore (1+Rf1/Ri1).
Finally, the output of the third stage Aout 150 is shown passing through a passive high-pass filter consisting Rout 171 and Cout 172. As with the stages Ain 110 and Aout 150, this element is not required for the function of the sample and hold circuit but is included for the specific example now described.
In this example, Vin is taken to have a peak voltage of from 0.5-1 volts in which is embedded a signal in the frequency range of 10 Hz to 1 kHz with an amplitude of from 25 μV to 75 μV. For simplicity, the circuit is taken to operate between ground and a Vdd of around 5V. The circuit needs to extract this embedded signal from, say, a microwave transducer or other source and amplify it sufficiently to be usable as an output, requiring a gain of around 90-103 dB.
The first stage Ain 110 acts as a buffer and reduces the gain requirements of the later stages. Its gain is give by (1+Rf1Ri1) and amplifies both the DC level and the embedded signal. This gain can either be variable or fixed depending on the application of the circuit. For the example, as the peak voltage is 0.5-1V and Vdd≈5V, a gain of 2-4 will place the input voltage in the middle of the operating range. Depending on implementation, values for Rf1≈2Ri1 of 21-63 kΩ are reasonable to keep power dissipation down. This amplified signal is then passed on to the sample and hold stage S/H 120.
In S/H 120, the signal first passes through Ri2 123 and then on to switches S1 125 and S2 126. These switches are preferable implemented using some type FET technology to minimize any offset biases. The are controlled by a signal Vsam and arranged so that in the sample mode S1 125 is on and S2 126 is off, while in the hold mode S2 126 is on and S1 125 is off. As described above, in this way the capacitor Chold 124 integrates the input voltage from Ain 110 during the hold stage and minimizes signal feedthrough during the hold stage. For the example, Vsam is taken to be a pulse 7 μs wide and to repeat every 250 μs, as shown in FIG. 3a. (The scale of the axes differ from both of FIGS. 3b and 3 c below.) This results in a sampling frequency several times higher than the upper end of the frequency range needing to be extracted and consequently above the corresponding Nyquist frequency. For the Vsam shown in FIG. 3a repeats every 250 μs, corresponding to a sampling frequency of 4 kMz. Depending on the type and frequency of the signal being extracted, the sampling frequency is chosen accordingly, but keeping the sample mode much shorter than the hold mode. Although the waveform for Vsam in FIG. 3a is shown with an amplitude of 1 V, this value is not critical so long as it is sufficient to operate the switches S1 125 and S2 126.
When in sample mode, the transfer function of S/H 120 is given by equation (A), but with Rf replaced by the parallel combination of Rf2 122 and Chold 124. In hold mode, the output of S/H is set by the charge integrated on Chold 124 due to the voltage input from Ain 110, but with the DC level of that section cancelled and referenced to the new baseline Vref1. Thus, if configured for unity gain, the hold output would just be Vin with the DC level set to Vref1. For the exemplary application, the values of Rf2≈400 kΩ, Ri2≈2 kΩ, Raz1≈5 MΩ, Caz1≈0.3 μF, and Chold≈25-100 pf give the appropriate filter characteristic over 10-1000 Hz and result in a gain of around 115 over this frequency range. More generally, Ri is taken to be above 1 kΩ or so to limit power dissipation and the relative value of Rf picked to set the amount gain desired. Raz and Caz can then be taken to adjust filter pole. To increase flexibility, Caz1 (and possibly also Raz1) can be off-chip allowing the user to set the filter characteristics according to the application. The value of Chold is taken as reasonable size so that leakage is not a concern, but small enough to not use up too much chip area.
A convenient choice for Vref1 is ½Vdd, or, more generally ½(Vhigh−Vlow) for a device operating between logic values Vhigh and Vlow, as this choice places the output in the center of the operating range to minimize the possibility of saturation on the extracted signal. This choice can itself be offset to compensate for offset voltages in the components of S/H 120. Vref1 can be set by a resistive voltage divider or other standard technique. In alternate embodiments it can be made adjustable if this feature is useful for later stages of the circuit.
The output of S/H 120 will then consist of the time varying portion of Vin, which has been extracted, amplified with a gain of 300-400, and re-referenced to the DC level Vref1. For the example, the smallness of the embedded signal requires more amplification and the third stage Aout 150 provides the additional gain.
The parameters in Aout 150 are here taken to be the same as the corresponding values in S/H 120 and for the same reasons: Rf3≈400 kΩ, Ri3≈2 kΩ, Raz2≈5 MΩ, and Caz2≈0.3 μF. This allows the required gain to be fairly evenly split between Aout 150 and S/H 120, each stage producing a gain of around 115 over the frequency range. Although this arrangement is often convenient, as the amplification duties are evenly split and it allows for duplication of components, in other applications it may be advantageous to set the parameters in Aout 150 differently form S/H 120. The differences due to the additional sample and hold function of S/H 120, and the introduction of the switches and Chold because of them, may require parameter changes in S/H 120 relative to Aout 150. For example, a particular application of the circuit of FIG. 2 may require a value of Chold 124 that would result in the other parameters S/H 120 needing to be adjusted with respect to those in Aout 150.
For the same reasons as given above with respect to Vref1, a convenient choice for many applications Vref2 is Vref2=½V/dd. However, depending what the ultimate use to which Vout will be put, Vref2 can be set to a different value from Vref1 at whatever reference level is appropriate. For example, a given application may need the extracted signal referenced to ground. In a more general embodiment, Vref2 is adjustable to make its output less application dependent.
Finally, the signal passes through Rout 171 and Cout 172. This acts as a low pass filter to skim off any high frequency noise. For the example, the values Rout≈1 kΩ and Cout≈0.1 μF are suitable. As with the auto-zero feedback loops, this capacitor and also possibly resistor can be off chip to increase flexibility.
Exemplary waveforms for Vout and Vin for these values are respectively shown in FIGS. 3cand 3 b. FIG. 3b shows a sine wave having a frequency of 50 Hz and amplitude of 20 μV embedded in a 0.6V DC voltage. (Note that the Vin axis begins at 599.90 mV.) The resultant Vout is shown in FIG. 3c as having the same frequency as the embedded sine wave, but with an amplitude of 0.9V and a DC level of 2.5V, the offset of Vin having been removed and set to the new reference level of ½Vdd.
The circuit of FIG. 2 can be readily implemented on a single integrated circuit since it does not require the sort of large, off-chip capacitors needed in the gain path of the prior art described in the Background section to obtain a high gain, high pass sample and hold system. This circuit can be implemented as a CMOS circuit or, more generally, in any of the other standard technologies (BiCMOS, GaAs, SiGe, and so on). In this way, the reliability problems associated with discrete solutions—board leakage around the sample and hold FET, failure of the large electrolytic capacitors in the gain path, susceptibility of these capacitors to induced electromagnetic fields—are avoided. Additionally, the described solution reduces both board area and the number of components and allows a manufacturing flow using only surface mount components.
In other embodiments, it may be preferable to place the integrating capacitors (Caz1 142 and Caz2 162 of FIG. 2) and possibly also the input resistors (Raz1 143 and Raz2 163) for the auto-zero feedback loop off-chip. In this way, these elements can be chosen to set the filter characteristics to the particular application where the circuit is used. Otherwise, the circuit is restricted to the filter profile set when it is manufactured. Also, for the example used above where the desired frequencies run from 10 Hz to 1 kHz, values of Caz≈0.3 μF and Raz≈5 MΩ are used. A capacitor of this size would require a large amount of chip area and is therefore generally placed off-chip. Similarly, for reasons of flexibility and reduced chip size, Cout 172 is preferably not included on the chip for embodiments having the auto-zero capacitors off-chip.
Although these last embodiments may result in off-chip capacitors, these differ in both a qualitative and quantitative way from those placed in the gain path as found in the prior art. To meet the requirements of the given example, where a gain of several orders of magnitude is required starting from as low as 10 Hz while still eliminating the DC component, the sort of prior art designs described in the background section utilize capacitors with values in the range of 10 μF. Using smaller values would require unreasonable resistor values and likely lead to circuit leakage problems. Also, to obtain a capacitance of that size, the capacitors take up considerable area on a printed circuit board and are generally tantilum/electrolytic devices. This sort of device is usually found to be less reliable due the chemical nature the dielectric material. In the embodiments of the previous paragraph, the values used for Caz1 142 and Caz2 162 are one or two orders of magnitude smaller and consequently avoid these problems.
Various details of the implementation and method are merely illustrative of the invention. It will be understood that various changes in such details may be within the scope of the invention, which is to be limited only by the appended claims.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US3696305||Jul 1, 1970||Oct 3, 1972||Gen Electric||High speed high accuracy sample and hold circuit|
|US4119960||Feb 11, 1977||Oct 10, 1978||Siliconix Incorporated||Method and apparatus for sampling and holding an analog input voltage which eliminates offset voltage error|
|US4302689||Aug 2, 1979||Nov 24, 1981||John Fluke Mfg. Co., Inc.||Sample and hold circuit|
|US4417160||Jul 30, 1981||Nov 22, 1983||Rca Corporation||Offset compensation apparatus for biasing an analog comparator|
|US4587443||Aug 27, 1984||May 6, 1986||Signetics Corporation||Auto-zero sample and hold circuit|
|US4962323||Jul 12, 1989||Oct 9, 1990||National Semiconductor Corporation||High speed auto zero comparator|
|US5177382||Nov 26, 1991||Jan 5, 1993||Toko, Inc.||Active filter circuit|
|US5182522||Apr 3, 1992||Jan 26, 1993||Toko, Inc.||Polar leapfrog filter|
|US5297424||Mar 7, 1990||Mar 29, 1994||Monroe Auto Equipment Company||Telemetry system for tire pressure and temperature sensing|
|US5648738 *||Nov 1, 1994||Jul 15, 1997||Cirrus Logic, Inc.||Read channel having auto-zeroing and offset compensation, and power-down between servo fields|
|US5977803 *||Aug 19, 1997||Nov 2, 1999||Mitsubishi Denki Kabushiki Kaisha||Capacitance type sensor interface circuit|
|US6175254 *||Jan 29, 1999||Jan 16, 2001||Rochester Microsystems, Inc.||System for compensating a signal for an offset from a reference level|
|USH1616||May 31, 1994||Dec 3, 1996||Minnesota Mining And Manufacturing Company||Web inspection system having enhanced video signal preprocessing|
|EP0273425A2||Dec 28, 1987||Jul 6, 1988||Sony Corporation||Filter circuit|
|GB2215544A||Title not available|
|1||Gasparik, F. (1980) "A Precision Autozeroing Sample and Hold Integrated Circuit" IEEE Journal of Solid-State Circuits, SC-15(6):945-949.|
|2||Gasparik, F. (1980) "Session XI: Data Acquisition Components -THAM 11.2: An Autozeroing Sample and Hold IC " 1980 IEEE Solid State Circuits Conference Digest of Technical Papers, pp. 132-133.|
|3||Harris, F. et al. (1989) "New Results with Oversampled Converters" Conference Record of the Twenty-third Asilomar Conference on Signals, Systems & Computers, 2:844-848.|
|4||Hoskins, K. (1985) "Active Offset Correction Integrated in CMOS Technology" Elektronik Industrie, 16(6):96, 98, 100 and 102 (English language abstract provided).|
|5||Hribik, J. (1999) "Computer-To-Frequency Converters with Feedback" Journal of Electrical Engineering, 50(1-2):44-46.|
|6||Ishikawa, M. et al. (1989) "An 8-bit 50-MHz CMOS Subranging A/D Converter with Pipelined Wide-Band S/H" IEEE Journal of Solid-State Circuits, 24(6):1485-1491.|
|7||Kudo, M. et al. (1986) "A Proposal of Various Recursive Switched Capactor Filters Using a Single OP-Amp and Experimental Results" Electronics and Communications in Japan, Part 2, 69(2):84-92.|
|8||Matej, V. (1990) "A Ramp Generator" Sdelovaci Technika, 38(6):224-225 (English language abstract provided).|
|9||Real, P. et al. (1991) "TPM 10:1: A 14b Linear, 250ns Sample-and-Hold Subsystem with Self-Correction" IEEE International Solid State Circuits Conference DIgest of Technical Papers, pp. 164-165.|
|10||Shovak, A. et al. (1992) "Median-Based Offset Cancellation Circuit Technique" 1992 IEEE International Symposium on Circuits and System, 4:2033-2036.|
|11||Signell, S. et al. (1989) "Offset-Compensatio of Two-Phase Switched-Capacitor Filters" IEEE Transactions on Circuits and Systems, 36(1):31-41.|
|12||Skribanowitz, G. (1990) Low Frequency Filter VF 30) Radio Fernsehen Elektronik, 39(91):35-37 (English language abstract provided).|
|13||Sone, K. et al. (1993) "A 10-b 100-Msample/s Pipelined Subranging BiCMOS ADC" IEEE Journal of Solid-State Circuits, 28(12):1180-1186.|
|14||Wiegerink, R. et al. (1989) 37 Offset cancelling Circuit) IEEE Journal of Solid-State Circuits, 24(3):651-658.|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US6657487 *||Feb 5, 2001||Dec 2, 2003||Em(Us) Design, Inc||Photodetector preamplifier circuit having a rotating input stage|
|US8223911||May 21, 2007||Jul 17, 2012||Deutsches Elektronen-Synchrotron Desy||Self-compensating drift-free high-frequency phase detector circuit|
|US8374304 *||Feb 12, 2013||Denso Corporation||Filter circuit for removing noise pulses from digital input signal|
|US8810311 *||Jul 25, 2012||Aug 19, 2014||Texas Instruments Incorporated||Auto-zeroed amplifier with low input leakage|
|US20080174338 *||Jan 22, 2008||Jul 24, 2008||Denso Corporation||Filter circuit for removing noise pulses from digital input signal|
|US20110122977 *||May 21, 2007||May 26, 2011||Deutsches Elektronen-Synchrotron Desy||Self-compensating drift-free high-frequency phase detector circuit|
|US20130241637 *||Jul 25, 2012||Sep 19, 2013||Texas Instruments Incorporated||Auto-Zeroed Amplifier with Low Input Leakage|
|U.S. Classification||327/94, 327/95|
|Aug 3, 2000||AS||Assignment|
Owner name: ZILOG, INC., CALIFORNIA
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:LEITH, JAMES W.;REEL/FRAME:010967/0721
Effective date: 20000612
|Jun 30, 2005||FPAY||Fee payment|
Year of fee payment: 4
|Jun 22, 2009||FPAY||Fee payment|
Year of fee payment: 8
|Sep 3, 2010||AS||Assignment|
Owner name: IXYS CH GMBH, SWITZERLAND
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:ZILOG, INC.;REEL/FRAME:024964/0132
Effective date: 20100217
|Jun 13, 2013||FPAY||Fee payment|
Year of fee payment: 12
|May 19, 2015||AS||Assignment|
Owner name: IXYS INTL LIMITED, CAYMAN ISLANDS
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:IXYS CH GMBH;REEL/FRAME:035665/0589
Effective date: 20150515